吾爱破解 - 52pojie.cn

 找回密码
 注册[Register]

QQ登录

只需一步,快速开始

查看: 2153|回复: 25
收起左侧

[Python 原创] 利用腾讯云实现PDF转“word”

  [复制链接]
lianxiang1122 发表于 2024-6-17 16:51
PDF转word很是让人头疼,特别是扫描版的又有图片又有表格的,基本很难搞定。

最近,刚刚发现腾讯云刚刚发布了一个文字识别新品—————智能结构化!

https://cloud.tencent.com/product/smart-ocr

进入网址后,用微信登录并开通服务就可免费使用,腾讯云一如既往的优秀,可以每月白嫖1千次!

image.png

点击demo就可以先体验一下了,只能一张一张PDF或图片识别,识别后会生成一个markdown文件,用相关的软件打开这个markdown文件就可以复制粘贴到word文档中了,从而间接实现PDF转word。
能力有限,不能一次到位,只能这样了。。。。。等高手大佬们解决一次到位转成word。。。

image.png

例如,用我们的pycharm软件打开.md文件后,会显示内容,我们全选复制,再粘贴到word中,就可以了。
image.png

当然了,一张一张识别肯定会很麻烦了。腾讯云提供了API调用,找到控制台的接入指引,第二步的AIP3.0 Explorer,点击进入
image.png

进入AIP3.0 Explorer,之前分享过,不一一解释了。

image.png

将代码复制到IDE中,就可以测试了。

提示:一次只能识别10页PDF,如果多于10页,就得多次调用了。

具体操作,就不一一详解了。利用chatgpt,做了一个带界面的,供大家参考。

简单解释一下:先读取PDF文件,再选择生成markdown文件保存位置,输入你的ID和KEY。程序会先读取D盘是否有一个idkey.txt文件,如果有自动读取到相关的ID和KEY,如果没有,需要你输入一下,当点击转换按钮后,会自动将ID 和key保存到D盘,当下一次使用时读取,就不用再输入了。当pdf文件大于10页后,拆分后拼接,多次请求,最后生成一个.md文件。

[Python] 纯文本查看 复制代码
import os
import tkinter as tk
from tkinter import filedialog, messagebox
import base64
import fitz  # PyMuPDF
import json
from tencentcloud.common.common_client import CommonClient
from tencentcloud.common import credential
from tencentcloud.common.profile.client_profile import ClientProfile
from tencentcloud.common.profile.http_profile import HttpProfile
from tencentcloud.common.exception.tencent_cloud_sdk_exception import TencentCloudSDKException

def read_pdf_to_base64(paf_file):
    with open(paf_file, 'rb') as pdf_file:
        binary_data = pdf_file.read()
    base64_encoded_data = base64.b64encode(binary_data)
    return base64_encoded_data.decode('utf-8')

def decode_base64_to_markdown(base64_str):
    decoded_bytes = base64.b64decode(base64_str)
    return decoded_bytes.decode('utf-8')

def save_as_md_file(content, filename):
    with open(filename, 'w', encoding='utf-8') as file:
        file.write(content)

def ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key):
    cred = credential.Credential(secret_id, secret_key)
    httpProfile = HttpProfile()
    httpProfile.endpoint = "ocr.tencentcloudapi.com"
    clientProfile = ClientProfile()
    clientProfile.httpProfile = httpProfile

    params_set = {
        "FileType": "PDF",
        "FileBase64": base64_string,
        "FileStartPageNumber": FileStartPageNumber,
        "FileEndPageNumber": FileEndPageNumber
    }

    params = json.dumps(params_set)

    common_client = CommonClient("ocr", "2018-11-19", cred, "ap-guangzhou", profile=clientProfile)

    try:
        response = common_client.call_json("ReconstructDocument", json.loads(params))
        return response['Response']['MarkdownBase64']
    except TencentCloudSDKException as err:
        error_message = f"An error occurred: {err}\nPlease enter the correct API Secret ID and Secret Key."
        messagebox.showerror("Error", error_message)
        
        idkey_file = 'D:/idkey.txt'
        if os.path.exists(idkey_file):
            if messagebox.askyesno("Delete idkey.txt", "The idkey.txt file exists. Do you want to delete it?"):
                os.remove(idkey_file)
                messagebox.showinfo("Deleted", "idkey.txt file has been deleted. Please enter the correct API Secret ID and Secret Key.")
        return None

def process_pdf(paf_file, secret_id, secret_key, output_dir):
    base64_string = read_pdf_to_base64(paf_file)
    doc = fitz.open(paf_file)
    page_count = doc.page_count

    markdown_output = ''
    pdf_filename = os.path.splitext(os.path.basename(paf_file))[0]
    output_filepath = os.path.join(output_dir, f"{pdf_filename}.md")

    if page_count <= 10:
        FileStartPageNumber = 1
        FileEndPageNumber = page_count
        output = ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key)
        if output:
            markdown_output = decode_base64_to_markdown(output)
            save_as_md_file(markdown_output, output_filepath)
    else:
        num = page_count // 10
        num_last = page_count % 10
        for i in range(num):
            FileStartPageNumber = i * 10 + 1
            FileEndPageNumber = i * 10 + 10
            output = ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key)
            if output:
                markdown_output += decode_base64_to_markdown(output) + '\n'
        if num_last != 0:
            FileStartPageNumber = num * 10 + 1
            FileEndPageNumber = num * 10 + num_last
            output = ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key)
            if output:
                markdown_output += decode_base64_to_markdown(output) + '\n'
        save_as_md_file(markdown_output, output_filepath)

def select_file():
    file_path = filedialog.askopenfilename(filetypes=[("PDF files", "*.pdf")])
    file_entry.delete(0, tk.END)
    file_entry.insert(0, file_path)

def select_output_dir():
    output_dir = filedialog.askdirectory()
    output_dir_entry.delete(0, tk.END)
    output_dir_entry.insert(0, output_dir)

def start_conversion():
    paf_file = file_entry.get()
    output_dir = output_dir_entry.get()

    # Check if idkey.txt exists and read the keys
    idkey_file = 'D:/idkey.txt'
    if os.path.exists(idkey_file):
        with open(idkey_file, 'r') as file:
            lines = file.readlines()
            secret_id = lines[0].strip().split(': ')[1]
            secret_key = lines[1].strip().split(': ')[1]
    else:
        secret_id = secret_id_entry.get()
        secret_key = secret_key_entry.get()
        
        if not secret_id or not secret_key:
            messagebox.showerror("Error", "Please enter the API Secret ID and Secret Key.")
            return
        
        with open(idkey_file, 'w') as file:
            file.write(f"API Secret ID: {secret_id}\n")
            file.write(f"API Secret Key: {secret_key}\n")

    if not secret_id or not secret_key:
        messagebox.showerror("Error", "Please enter the API Secret ID and Secret Key.")
        return

    process_pdf(paf_file, secret_id, secret_key, output_dir)

# 创建主窗口
root = tk.Tk()
root.title("PDF to Markdown Converter")

# 创建并放置标签和文本框
tk.Label(root, text="Select PDF File:").grid(row=0, column=0, padx=10, pady=10)
file_entry = tk.Entry(root, width=50)
file_entry.grid(row=0, column=1, padx=10, pady=10)
tk.Button(root, text="Browse", command=select_file).grid(row=0, column=2, padx=10, pady=10)

tk.Label(root, text="Select Output Directory:").grid(row=1, column=0, padx=10, pady=10)
output_dir_entry = tk.Entry(root, width=50)
output_dir_entry.grid(row=1, column=1, padx=10, pady=10)
tk.Button(root, text="Browse", command=select_output_dir).grid(row=1, column=2, padx=10, pady=10)

tk.Label(root, text="API Secret ID:").grid(row=2, column=0, padx=10, pady=10)
secret_id_entry = tk.Entry(root, width=50)
secret_id_entry.grid(row=2, column=1, padx=10, pady=10)

tk.Label(root, text="API Secret Key:").grid(row=3, column=0, padx=10, pady=10)
secret_key_entry = tk.Entry(root, width=50, show='*')
secret_key_entry.grid(row=3, column=1, padx=10, pady=10)

tk.Button(root, text="Convert", command=start_conversion).grid(row=4, column=1, pady=20)

# 运行主循环
root.mainloop()

免费评分

参与人数 5吾爱币 +11 热心值 +5 收起 理由
qq63 + 1 + 1 用心讨论,共获提升!
ddddx + 1 + 1 我很赞同!
苏紫方璇 + 7 + 1 欢迎分析讨论交流,吾爱破解论坛有你更精彩!
丶峰宇 + 1 + 1 用心讨论,共获提升!
莫奇 + 1 + 1 谢谢 @Thanks!

查看全部评分

本帖被以下淘专辑推荐:

发帖前要善用论坛搜索功能,那里可能会有你要找的答案或者已经有人发布过相同内容了,请勿重复发帖。

 楼主| lianxiang1122 发表于 2024-8-15 11:44
有段时间没有用这个了,要不是楼上qq63还真没发现这货居然修改了服务器。。。。。。调到“ 文档解析”里去了。。。我说怎么一提交就报错呢。。。。

参考官方说明:https://cloud.tencent.com/document/api/866/104610

主要修改如下:
1.
[Python] 纯文本查看 复制代码
#httpProfile.endpoint = "ocr.tencentcloudapi.com" 之前
    httpProfile.endpoint = "lke.tencentcloudapi.com"  #修改后


2.
[Python] 纯文本查看 复制代码
#"FileType": "PDF", 取消pdf设置,直接删除改行代码


3.

[Python] 纯文本查看 复制代码
#common_client = CommonClient("ocr", "2018-11-19", cred, "ap-guangzhou", profile=clientProfile)  之前
    common_client = CommonClient("lke", "2023-11-30", cred, "ap-guangzhou", profile=clientProfile)   #现在


完整代码修改参考如下:

[Python] 纯文本查看 复制代码
import os
import tkinter as tk
from tkinter import filedialog, messagebox
import base64
import fitz  # PyMuPDF
import json
from tencentcloud.common.common_client import CommonClient
from tencentcloud.common import credential
from tencentcloud.common.profile.client_profile import ClientProfile
from tencentcloud.common.profile.http_profile import HttpProfile
from tencentcloud.common.exception.tencent_cloud_sdk_exception import TencentCloudSDKException

def read_pdf_to_base64(paf_file):
    with open(paf_file, 'rb') as pdf_file:
        binary_data = pdf_file.read()
    base64_encoded_data = base64.b64encode(binary_data)
    return base64_encoded_data.decode('utf-8')

def decode_base64_to_markdown(base64_str):
    decoded_bytes = base64.b64decode(base64_str)
    return decoded_bytes.decode('utf-8')

def save_as_md_file(content, filename):
    with open(filename, 'w', encoding='utf-8') as file:
        file.write(content)

def ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key):
    cred = credential.Credential(secret_id, secret_key)
    # 实例化一个http选项,可选的,没有特殊需求可以跳过
    httpProfile = HttpProfile()
    httpProfile.endpoint = "lke.tencentcloudapi.com"
    clientProfile = ClientProfile()
    clientProfile.httpProfile = httpProfile

    params_set = {
        #"FileType": "PDF",
        "FileBase64": base64_string,
        "FileStartPageNumber": FileStartPageNumber,
        "FileEndPageNumber": FileEndPageNumber
    }

    params = json.dumps(params_set)

    common_client = CommonClient("lke", "2023-11-30", cred, "ap-guangzhou", profile=clientProfile)

    try:
        response = common_client.call_json("ReconstructDocument", json.loads(params))
        return response['Response']['MarkdownBase64']
    except TencentCloudSDKException as err:
        error_message = f"An error occurred: {err}\nPlease enter the correct API Secret ID and Secret Key."
        messagebox.showerror("Error", error_message)
        
        idkey_file = 'D:/idkey.txt'
        if os.path.exists(idkey_file):
            if messagebox.askyesno("Delete idkey.txt", "The idkey.txt file exists. Do you want to delete it?"):
                os.remove(idkey_file)
                messagebox.showinfo("Deleted", "idkey.txt file has been deleted. Please enter the correct API Secret ID and Secret Key.")
        return None

def process_pdf(paf_file, secret_id, secret_key, output_dir):
    base64_string = read_pdf_to_base64(paf_file)
    doc = fitz.open(paf_file)
    page_count = doc.page_count

    markdown_output = ''
    pdf_filename = os.path.splitext(os.path.basename(paf_file))[0]
    output_filepath = os.path.join(output_dir, f"{pdf_filename}.md")

    if page_count <= 10:
        FileStartPageNumber = 1
        FileEndPageNumber = page_count
        output = ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key)
        if output:
            markdown_output = decode_base64_to_markdown(output)
            save_as_md_file(markdown_output, output_filepath)
    else:
        num = page_count // 10
        num_last = page_count % 10
        for i in range(num):
            FileStartPageNumber = i * 10 + 1
            FileEndPageNumber = i * 10 + 10
            output = ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key)
            if output:
                markdown_output += decode_base64_to_markdown(output) + '\n'
        if num_last != 0:
            FileStartPageNumber = num * 10 + 1
            FileEndPageNumber = num * 10 + num_last
            output = ocr_markdown(base64_string, FileStartPageNumber, FileEndPageNumber, secret_id, secret_key)
            if output:
                markdown_output += decode_base64_to_markdown(output) + '\n'
        save_as_md_file(markdown_output, output_filepath)

def select_file():
    file_path = filedialog.askopenfilename(filetypes=[("PDF files", "*.pdf")])
    file_entry.delete(0, tk.END)
    file_entry.insert(0, file_path)

def select_output_dir():
    output_dir = filedialog.askdirectory()
    output_dir_entry.delete(0, tk.END)
    output_dir_entry.insert(0, output_dir)

def start_conversion():
    paf_file = file_entry.get()
    output_dir = output_dir_entry.get()

    # Check if idkey.txt exists and read the keys
    idkey_file = 'D:/idkey.txt'
    if os.path.exists(idkey_file):
        with open(idkey_file, 'r') as file:
            lines = file.readlines()
            secret_id = lines[0].strip().split(': ')[1]
            secret_key = lines[1].strip().split(': ')[1]
    else:
        secret_id = secret_id_entry.get()
        secret_key = secret_key_entry.get()
        
        if not secret_id or not secret_key:
            messagebox.showerror("Error", "Please enter the API Secret ID and Secret Key.")
            return
        
        with open(idkey_file, 'w') as file:
            file.write(f"API Secret ID: {secret_id}\n")
            file.write(f"API Secret Key: {secret_key}\n")

    if not secret_id or not secret_key:
        messagebox.showerror("Error", "Please enter the API Secret ID and Secret Key.")
        return

    process_pdf(paf_file, secret_id, secret_key, output_dir)

# 创建主窗口
root = tk.Tk()
root.title("PDF to Markdown Converter")

# 创建并放置标签和文本框
tk.Label(root, text="Select PDF File:").grid(row=0, column=0, padx=10, pady=10)
file_entry = tk.Entry(root, width=50)
file_entry.grid(row=0, column=1, padx=10, pady=10)
tk.Button(root, text="Browse", command=select_file).grid(row=0, column=2, padx=10, pady=10)

tk.Label(root, text="Select Output Directory:").grid(row=1, column=0, padx=10, pady=10)
output_dir_entry = tk.Entry(root, width=50)
output_dir_entry.grid(row=1, column=1, padx=10, pady=10)
tk.Button(root, text="Browse", command=select_output_dir).grid(row=1, column=2, padx=10, pady=10)

tk.Label(root, text="API Secret ID:").grid(row=2, column=0, padx=10, pady=10)
secret_id_entry = tk.Entry(root, width=50)
secret_id_entry.grid(row=2, column=1, padx=10, pady=10)

tk.Label(root, text="API Secret Key:").grid(row=3, column=0, padx=10, pady=10)
secret_key_entry = tk.Entry(root, width=50, show='*')
secret_key_entry.grid(row=3, column=1, padx=10, pady=10)

tk.Button(root, text="Convert", command=start_conversion).grid(row=4, column=1, pady=20)

# 运行主循环
root.mainloop()
 楼主| lianxiang1122 发表于 2024-11-26 14:42
哎,这货又改了。。。。。。。不如以前用的爽了。。。。改到大模型知识引擎 LKE的原子能力里了。。。没有一个月1000次的免费额度了,改为试用有效期2个月,免费额度大概50万tokens了,哎,不能好好的玩耍了。。。临时需要的可以去以下网址开通:

https://cloud.tencent.com/product/lke

然后去

https://cloud.tencent.com/document/product/1759/107504,

找到点击调试,进入 API Explorer调试界面获取相关的代码,跟之前的用法一样。

屏幕截图 2024-11-26 144110.png
huazhongxu 发表于 2024-6-17 17:06
dayday26 发表于 2024-6-17 17:18
感谢分享,不知道能不能完美实现表格转换
ZhjhJZ 发表于 2024-6-17 17:26
腾讯云一如既往的优秀,可以每月白嫖1千次!牛!
饮食协会 发表于 2024-6-17 17:49
带界面在哪里?
bluepeb 发表于 2024-6-17 19:01
综合各种技术应用,全方位的高手!
lvtaode0657 发表于 2024-6-17 21:18
好东西啊,谢谢分享
Psc7day 发表于 2024-6-18 00:14
感谢分享
kidll 发表于 2024-6-18 09:11
我52破解的坤,渍渍渍
Vincent168 发表于 2024-6-18 09:28
感谢楼主分享
您需要登录后才可以回帖 登录 | 注册[Register]

本版积分规则

返回列表

RSS订阅|小黑屋|处罚记录|联系我们|吾爱破解 - LCG - LSG ( 京ICP备16042023号 | 京公网安备 11010502030087号 )

GMT+8, 2024-12-13 01:15

Powered by Discuz!

Copyright © 2001-2020, Tencent Cloud.

快速回复 返回顶部 返回列表